Well from my testing...
1.PCI-e drives improves sql db reads/writes faster
2. Terminal server Epicor app runs faster than desktop Epicor app
3. 25-30 users per appserver using hardware load balancer
4. SSRS reports gnenrate faster than crystal
1.PCI-e drives improves sql db reads/writes faster
2. Terminal server Epicor app runs faster than desktop Epicor app
3. 25-30 users per appserver using hardware load balancer
4. SSRS reports gnenrate faster than crystal
--- In vantage@..., "s1mhall" <s1mhall@...> wrote:
>
> Looking forward to V10, but not the pain in between times. We still have our CRM, and support teams to migrate. Then work on adding all the rest such as APM plus, EPM, and Portal......I'm tired just thinking about it.
>
> S
>
> --- In vantage@yahoogroups.com, xorone <xorone@> wrote:
> >
> > With regards to your SSD references.  They DO give Vantage a significant performance boost.
> >
> > Chris
> >
> >
> > ________________________________
> > From: "Winter, Patrick" <pjw@>
> > To: vantage@yahoogroups.com
> > Sent: Thursday, June 13, 2013 1:30 PM
> > Subject: RE: [Vantage] Re: Separating AppServers and Database
> >
> >
> >
> > ÂÂ
> > Epicor would probably reply with that is exactly what we have done in V10 with a clean stack on SQL only.
> >
> >
> >
> > Patrick Winter
> >
> >
> >
> > From: vantage@yahoogroups.com [mailto:vantage@yahoogroups.com] On Behalf Of Chris Thompson
> > Sent: Thursday, June 13, 2013 12:05
> > To: vantage@yahoogroups.com
> > Subject: Re: [Vantage] Re: Separating AppServers and Database
> >
> >
> >
> >
> >
> > Does no one else think Epicor are just taking advantage to make their poorly written software run after instead of fixing it in the first place?
> >
> > Chris Thompson
> >
> > On 13 Jun 2013, at 17:07, "laraines422" <laraines422@ <mailto:laraines422%40yahoo.com> > wrote:
> >
> > > Simon:
> > >
> > > Thanks for the post. We're starting our upgrade project to E9 - and we really debated about the hardware specifications. We finally decided to bite the bullet and purchase SSD's for the new server.
> > >
> > > Your post helps to validate our decision - thank you!
> > >
> > > Laraine
> > >
> > > --- In vantage@yahoogroups.com <mailto:vantage%40yahoogroups.com> , "s1mhall" <s1mhall@> wrote:
> > > >
> > > > I have to agree with Jose here. We burnt 6 months of our implementation project with tooing and froeing with regards to the hardware setup. We initially had everything virtualized with SQL on a separate host and a dedicated 10Gb link between the host servers. We had a 4 node NLB cluster. Performance was average and we were seeing our performance results in the high end of the acceptable range based on the Epicor Performance tuning guide.
> > > >
> > > > If I were to go down the NLB route again then I'd be investigating F5 load balancers and hang the expense.
> > > >
> > > > Unfortunately the guide does not give any benchmark statistics on time entry, which would have been very helpful.
> > > >
> > > > Convinced that an upgrade to 701 would resolve the issues we went down that path with very little improvement. Finally after a hardware review conducted with Epicor we ditched Hyper-v and took one of the hosts and installed SQL 2008 R2, Progress Open Edge and Epicor 9. We have seen a marginal improvement.
> > > >
> > > > Submitting time can take from anywhere to 2-4 seconds on a good day and if you are a remote user then you are destined to wait longer (compression is enabled for remote users and that has helped). Epicor Web Access is used only for users who have a Mac and we may even migrate them to use the rich client on parallels. EWA is slower that the rich client to remote sites. Performing a Build Project Analysis (key step in project billing) takes about 4 minutes per project! Apparently there are some Build Project Analysis performance improvements in 702.
> > > >
> > > > Performance wise what we have seen is that using External BAQs where possible gives you a huge boost. We have performed some simple comparisons, although not flushing the SQL procedure cache etc. between tests, but complex BAQs certainly do benefit from building a view in SQL first. Just a shame you can't do it for stored procedures as well.
> > > >
> > > > On a further note about our time testing performance around submitting time. Testing using the education database, our Tech support person at Epicor was running rings around our performance stats and he was running E9 on a notebook with 2 SSDs! We have a 4 CPU 10 Core server with 196GB of RAM with 6 SAS disks in RAID 10! Other consultants I have discussed performance issues with, say they have seen some huge improvements using Fusion I/O cards, nice if you have the budget.
> > > >
> > > > As Jose says, in most instances there is very little gain from tweaking the environment and you can waste a lot of time reconfiguring. I agree, been there done that, ouch!
> > > >
> > > > Get the right hardware, get SSDs and be done with it! All of the users will thank you for it. If you don't have the budget then perhaps it's time to turn E9 into shelfware and go back to market.
> > > >
> > > > Epicor do have some white papers about performance (search epicweb for performance white paper). The early one around 2011 includes all the details as to what equipment they used and how they went about it. Please make sure you are sitting down when you read the hardware specs. Apparently it was done in a Microsoft Lab using Visual Studio tools to produce simulated transactional loads. The latter one focuses on performance improvements between versions. I have not checked recently if they have done an update.
> > > >
> > > > Apologies for the long rant er... post.
> > > >
> > > > Cheers
> > > > Simon
> > > >
> > > >
> > > >
> > > >
> > > > --- In vantage@yahoogroups.com <mailto:vantage%40yahoogroups.com> , "jseeman21" <jseeman@> wrote:
> > > > >
> > > > > Hello,
> > > > >
> > > > > We are currently running Epicor 905.701 and though we see some performance improvements from earlier versions, we are still looking to achieve better performance.
> > > > >
> > > > > It was suggested that we separate our appservers from the database. I feel silly asking this but can someone provide instruction on how to do this? We are currently running a Progress 64Bit environment. The entire epicor905 directory resides on a separate raid 10.
> > > > >
> > > > > In addition, we are also looking at using SSD drives. It was mentioned you can separate out just the DB onto another drive. Does anyone have instruction on how to isolate the epicor905\db directory?
> > > > >
> > > > > Thank you in advance for any assistance.
> > > > >
> > > > > Cheers.
> > > > > Jill Seeman
> > > > >
> > > >
> > >
> > >
> >
> > [Non-text portions of this message have been removed]
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>